MicroBooNE TPC Calibrations - Nevis Laboratoriesthe rest of the REU students Andrew Smith (Nevis Labs, Columbia) MicroBooNE TPC Calibrations 8/1/2014 31 / 36 Backup Slides Andrew Smith

MicroBooNE TPC Calibrations

Andrew Smith

Nevis Labs, Columbia

8/1/2014

Andrew Smith (Nevis Labs, Columbia) MicroBooNE TPC Calibrations 8/1/2014 1 / 36

Outline

1 MicroBooNE ExperimentMiniBooNEDetector

2 Electronics

3 CalibrationsGain/Linearity TestsCross TalkPeak Timing Test

4 Silicon Photomultipliers


MicroBooNE Experiment

Goals of MicroBooNE Experiment

1 Explore low energy neutrino interactions, where MiniBoone saw anunexplained excess

2 Further measure neutrino interaction cross sections

3 Test liquid argon detector


MiniBooNE

LSND, MiniBoone excessof νe

Different ∆m suggests4th masseigenstate→sterileneutrino

MicroBooNE created toexplore 0.1-1GeV area ofneutrino interactions

MiniBooNE cannotdiscern γ from electrons


MicroBooNE Detector

Liquid Argon acts as ionization medium as well as scintillator

170 ton cryostat

3 wire planes offset by 60 degrees, 2 induction, one collectionCharged particles drift toward wires and are collected8256 total wires

Photons are collected by PMTs behind wire planes


Liquid Argon

Figure: Cross section of MicroBooNEdetector

Figure: Example LArTPC, not actualMicroBooNE detector


Booster Neutrino Beam


Electronics


Electronics

Path of electronic signal

Wires ASIC ADC FPGA PC

1 Wires (or capacitor in parallel)- Inside detector

2 ASIC - Preamplifier, shapes pulse with an optimal time of 1µs

3 ADC - digitizes symbol with 12 bit range, at a frequency of 2 MHz

4 FPGA - processes and reduces data to reasonable amount and storesit temporarily

5 DAQ PC (Data Acquisition)- stores data for future analysis


Electronics


Calibrations

Test electronics readout each time the detector is moved or changedin any way

Write code modules to do analysis of the output data for when thedetector starts taking data

Figure: Example pulsed channel


Generation of Signal

1 Input voltage on capacitor induces charge on the wire

2 Signal goes through ASIC, with specific gain and shaping

There are 8256 waveforms created, one for each channel. For gain tests,the maximum ADC count is extracted from each channel’s waveform.


Gain/Linearity Tests

For each channel, fit the data to a line–get the gain from the slope

16 channels per asic

4 asics per FEM, 12 FEMs per feedthrough

Gain Calculations

C ∗VASIC = Qin ∗GASIC ∗Gint ∗GADC

Slope = GASIC ∗ Gint ∗ GADC/C


Linearity

Figure: Overlaid Linearity Curves for 768channels

4 gain settings, lower gain haslower slope

Saturation of ASIC

Induction wires have baseline of2000 ADCs

Collection wires baseline 450ADCs

One visible bad channel


Gain for Each Channel

Figure: One channel is probably bad, see extra large gainAndrew Smith (Nevis Labs, Columbia) MicroBooNE TPC Calibrations 8/1/2014 15 / 36

Cross Talk Analysis

For each subrun in the cross talk analysis:

Pulsed some channels, but not all

Look for pulses on unpulsed channels


Cross Talk

Looking at the Maximum ADC count for each channel shows that 1channel per ASIC was pulsed.


Cross Talk–Within ASIC

Cross Talk Formula

Cross Talk =MaxADC-Baseline for unpulsed channel

MaxADC-Baseline for pulsed channel

Figure: Empty Event: Cross talk levelsaround 1% Figure: Cross talk levels elevated for

induction channels


Cross Talk

RMS Noise

RMS =

√Σ(ADC − Baseline)2

N


Cross Talk

Empty Event Cross Talk Event


Cross Talk

Average Cross Talk over 100 events


Cross Talk Comparison

In a previous calibrations run, we pulsed a total of 192 channels instead of48. We saw the cross talk on induction channels increase linearly.

Figure: Pre-Move Figure: Post-Move


Cross Talk Conclusions

In the 2 data sets available (pre- and post-move), the change in crosstalk % is linearly related to the number of channels pulsed.

went from 0.6% cross talk to 2.5% cross talk when pulsing 4x thenumber of channels

Would eventually like to pulse 2, 3, 4, etc channels per ASIC to getmore complete data

Lack of map of channel number to actual wire proximity → no way totest cross talk vs proximity to wire

Most importantly, no new broken channels


Peak Timing Test

The trigger is delayed, and we want to look at the effects upon the pulse.For a single channel, took the average waveform over the entire subrun(100 events).


Peak Timing Test

The maximum ADC count for each subrun-delayed in 50ns increments


Peak Timing Test

Then calculated the integral over the pulse area for each subrun andcompared them based on the trigger delay.


Silicon Photomultipliers (SiPMs)

SiPMs are much smaller than PMTs

Lower gain than PMTs, but lower voltage required

Not sensitive to magnetic fields


SiPM Setup

1 Pulser sends trigger to LED

2 LED sends photons at SiPM

3 SiPM photons cause avalanche, amplification of signal

4 Signal goes to amplifier

5 View signal on oscilloscope


SiPMs

Looked at bias voltage of SiPM vs output


SiPMs

Looking to see single photon on oscilloscope. Noise levels at room tempmay be too high


Acknowledgments

A hearty thank you to everyone who helped to make this possible.

Mike Shaevitz

Leslie Camilleri

David Caratelli

Georgia Karagiorgi

David Kaleko

John Parsons

the NSF

the rest of the REU students


Backup Slides


Run Data

Each run is organized in a particular fashion:

Each run contains 125 subruns, each of which contains 100 events

Each subrun is different and is used to test the electronics fordifferent gain and shaping settings

There are 20 subruns for peak timing tests, 15 for cross talk, etc

Each event is 100 time ticks at 50 Hz trigger rate

There is also a second run of long events where the noise is tested


Fit Residuals

Overlay of fit residuals for one feedthrough (64 channels). Some patternsuggests underlying nonlinearity.Andrew Smith (Nevis Labs, Columbia) MicroBooNE TPC Calibrations 8/1/2014 34 / 36

Cross Talk

Empty Event Cross Talk Event


Cross Talk RMS Noise

Figure: Pre-Move RMS NoiseDistribution

Figure: Post-move RMS NoiseDistribution


MicroBooNE ExperimentMiniBooNEDetector

ElectronicsCalibrationsGain/Linearity TestsCross TalkPeak Timing Test

Silicon Photomultipliers

Documents

MicroBooNE TPC Calibrations - Nevis Laboratoriesthe rest of the REU students Andrew Smith (Nevis Labs, Columbia) MicroBooNE TPC Calibrations 8/1/2014 31 / 36 Backup Slides Andrew Smith